The next generation lithographies (NGL) for advanced computer chip manufacturing have required the development of technologies such as extreme ultraviolet lithography(EUVL) as a potential solution. This lithographic approach generally relies on the use of multiplayer-coated reflective optics that has narrow pass bands in a spectral region where conventional transmissive optics is inoperable. Laser plasmas and electric discharge type plasmas are now considered prime candidate sources for the development of EUV. The requirements of this source, in output performance, stability and operational life are considered extremely stringent. At the present time, the wavelengths of choice are approximately 13 nm and 11.7 nm. This type of source must comprise a compact high repetition rate laser and a renewable target system that is capable of operating for prolonged periods of time. For example, a production line facility would require uninterrupted system operations of up to three months or more. That would require an uninterrupted operation for some 10 to the 9th shots, and would require the unit shot material costs to be in the vicinity of 10 to minus 6 so that a full size stepper can run at approximately 40 to approximately 80 wafer levels per hour. These operating parameters stretch the limitations of conventional laser plasma facilities.
Generally, laser plasmas are created by high power pulsed lasers, focused to micron dimensions onto various types of solids or quasi-solid targets, that all have inherent problems. For example, U.S. Pat. No. 5,151,928 to Hirose described the use of film type solid target tapes as a target source. However, these tape driven targets are difficult to construct, prone to breakage, costly and cumbersome to use and are known to produce low velocity debris that can damage optical components such as the mirrors that normally used in laser systems.
Other known solid target sources have included rotating wheels of solid materials such as Sn or tin or copper or gold, etc. However, similar and worse than to the tape targets, these solid materials have also been known to produce various ballistic particles sized debris that can emanate from the plasma in many directions that can seriously damage the laser system's optical components. Additionally these sources have a low conversion efficiency of laser light to in-band EUV light at only 1 to 3%.
Solid Zinc and Copper particles such as solid discs of compacted materials have also been reported for short wavelength optical emissions. See for example, T. P. Donaldson et al. Soft X-ray Spectroscopy of Laser-produced Plasmas, J. Physics, B:Atom. Molec. Phys., Vol. 9, No. 10. 1976, pages 1645–1655. FIGS. 1A and 1B show spectra emissions of solid Copper(Cu) and Zinc(Zn) targets respectively described in this reference. However, this reference requires the use of solid targets that have problems such as the generation of high velocity micro type projectiles that causes damage to surrounding optics and components. For example, page 1649, lines 33–34, of this reference states that a “sheet of mylar . . . was placed between the lens and target in order to prevent damage from ejected target material . . . .” Thus, similar to the problems of the previously identified solids, solid Copper and solid Zinc targets also produce destructive debris when being used. Shields such as mylar, or other thin film protectors may be used to shield against debris for sources in the X-ray range, though at the expense of rigidity and source efficiency. However, such shields cannot be used at all at longer wavelengths in the XUV and EUV regions.
Frozen gases such as Krypton, Xenon and Argon have also been tried as target sources with very little success. Besides the exorbitant cost required for containment, these gases are considered quite expensive and would have a continuous high repetition rate that would cost significantly greater than $10 to the minus 6. Additionally, the frozen gasses have been known to also produce destructive debris as well, and also have a low conversion efficiency factor.
An inventor of the subject invention previously developed water laser plasma point sources where frozen droplets of water became the target point sources. See U.S. Pat. Nos. 5,459,771 and 5,577,091 both to Richardson et al., which are both incorporated by reference. It was demonstrated in these patents that oxygen was a suitable emitter for line radiation at approximately 11.6 nm and approximately 13 nm. Here, the lateral size of the target was reduced down to the laser focus size, which minimized the amount of matter participating in the laser matter interaction process. The droplets are produced by a liquid droplet injector, which produces a stream of droplets that may freeze by evaporation in the vacuum chamber. Unused frozen droplets are collected by a cryogenic retrieval system, allowing reuse of the target material. However, this source displays a similar low conversion efficiency to other sources of less than approximately 1% so that the size and cost of the laser required for a full size 300 mm stepper running at approximately 40 to approximately 80 wafer levels per hour would be a considerable impediment.
Other proposed systems have included jet nozzles to form gas sprays having small sized particles contained therein, and jet liquids. See for Example, U.S. Pat. Nos. 6,002,744 to Hertz et al. and 5,991,360 to Matsui et al. However, these jets use many particles that are not well defined, and the use of jets creates other problems such as control and point source interaction efficiency. U.S. Pat. No. 5,577,092 to Kulak describe cluster target sources using rare expensive gases such as Xenon would be needed.
Attempts have been made to use a solid liquid target material as a series of discontinuous droplets. See U.S. Pat. No. 4,723,262 to Noda et al. However, this reference states that liquid target material is limited by example to single liquids such as “preferably mercury”, abstract. Furthermore, Noda states that “. . . although mercury as been described as the preferred liquid metal target, any metal with a low melting point under 100 C. can be used as the liquid metal target provided an appropriate heating source is applied. Any one of the group of indium, gallium, cesium or potassium at an elevated temperature may be used . . . ”, column 6, lines 12–19. Thus, this patent again is limited to single metal materials and requires an “appropriate heating source (be) applied . . . ” for materials other than mercury.